Stefan Zehle, CEO at Coleago Consulting, discusses the current spectrum auction landscape in Europe.

Eurocomms.com: You recently wrote a report on the outcomes of the 4G spectrum auctions in Europe – what did you find out?

Stefan Zehle: We saw vastly different prices in the 800MHz spectrum as a result of a number of different factors. Supply and demand was the most important, so we saw very high prices where the number of bidders exceeded the available spectrum. In Germany, for example, four bidders competed for 2x30MHz of 800MHz spectrum and prices were driven up until the weakest bidders decided to withdraw. On average in Europe, prices paid for 800MHz spectrum have been 15 times higher than for 2.6GHz spectrum due to the better propagation characteristics of the 800Mhz spectrum.

What other factors influenced the auctions?

It was interesting to note how the economic position of each country became a driving factor behind high reserve prices. The reserve price in France for the 2.6GHz spectrum, for example, was 25 times higher than that of Germany. Cash-strapped governments are looking to raise revenue and view high reserve prices as one way of doing that – what SFR paid reflected the reserve price, but in Germany operators only paid a quarter of the reserve price.

How does auction design impact the outcome and what advice can you offer operators?

Naturally, auction design has a lot to do with how much you end up paying so it can put operators in the driving seat to influence the rules. The consultation stage is one of the most important parts, so don’t leave it until the rules have been published – try to get involved before to get the right sort of design. You need to engage really early, so if an auction is due to happen in 2013 then you need to start thinking about this now.

Certain auction designs, such as “combinatorial clock auction” with a “second price rule”, result in surprising outcomes with some bidders paying much more per MHz of spectrum than others. The 2.6GHz auction in Denmark is a good example – three bidders paid 45 times more than the fourth bidder, Hutchison 3.

What auction design has worked best?

There is no “best-fit” model, it is very much market specific. For some regulators the primary concern is to maintain competition at the network level, while in digital dividend auctions mobile broadband coverage in rural areas is an objective. For other regulators raising cash is the main driver. Operators should try to persuade regulators to set low reserve prices to prevent it becoming a tax on the mobile industry.

It can be strongly argued that the benefits generated from deploying mobile broadband are a much more valuable form of revenue generation than simply making money from an auction. In this context Ireland is an interesting example because the regulator is looking to maximise auction proceeds. One of the main operators, Meteor, is owned by eircom, the incumbent fixed operator. Eircom is struggling with insolvency and it will be interesting to see how this will affect bidding in the auction to renew existing spectrum and bid for the 800MHz digital dividend spectrum.

Overall, what are the main lessons that can be learned from those auctions that have taken place?

For anyone going into future auctions we would highly encourage participating in the consultation process to get the best possible auction design and lowest reserve prices, value the spectrum to set bid limits and be prepared to execute the auction effectively to the bid strategy. The prices paid for 2.6GHz lots tend to be higher when sold in single-band auctions prior to the issuing of 800MHz, and it’s clear that there is enough 2.6GHz spectrum to grant each mobile operator the widest LTE channel width (2x20MHz) in FDD.